Differential Regulation of TLR-Dependent MyD88 and TRIF Signaling Pathways by Free Zinc Ions

This information is current as Anne Brieger, Lothar Rink and Hajo Haase of October 6, 2021. J Immunol 2013; 191:1808-1817; Prepublished online 17 July 2013; doi: 10.4049/jimmunol.1301261 http://www.jimmunol.org/content/191/4/1808 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2013/07/12/jimmunol.130126 Material 1.DC1

References This article cites 48 articles, 13 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/191/4/1808.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on October 6, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Differential Regulation of TLR-Dependent MyD88 and TRIF Signaling Pathways by Free Zinc Ions

Anne Brieger, Lothar Rink, and Hajo Haase

Zinc signals are utilized by several immune cell receptors. One is TLR4, which causes an increase of free zinc ions (Zn2+) that is required for the MyD88-dependent expression of inflammatory . This study investigates the role of Zn2+ on Toll/IL-1R domain–containing adapter inducing IFN-b (TRIF)–dependent signals, the other major intracellular pathway activated by TLR4. Chelation of Zn2+ with the membrane-permeable chelator N,N,N’,N’-Tetrakis(2-pyridylmethyl)ethylenediamine augmented TLR4-mediated production of IFN-b and subsequent synthesis of inducible NO synthase and production of NO. The effect is based on Zn2+ acting as a negative regulator of the TRIF pathway via reducing IFN regulatory factor 3 activation. This was also observed with TLR3, the only TLR that signals exclusively via TRIF, but not MyD88, and does not trigger a zinc signal. In contrast, IFN-g–induced NO production was unaffected by N,N,N’,N’-Tetrakis(2-pyridylmethyl)ethylenediamine. Taken together,

Zn2+ is specifically involved in TLR signaling, where it differentially regulates MyD88 and TRIF signaling via a zinc signal or via Downloaded from basal Zn2+ levels, respectively. The Journal of Immunology, 2013, 191: 1808–1817.

he essential trace element zinc is indispensable for the NF-kB activation, and thereby for transcription and release of function of the immune system. Zinc deficiency leads to inflammatory cytokines, such as TNF-a, IL-1b, and IL-6 (7). T increased susceptibility to infections, and restoring zinc After activation of the MyD88-dependent pathway, the levels of marginally deficient elderly by supplementation can im- complex is endocytosed (10) and the adaptors TRAM and Toll/IL- http://www.jimmunol.org/ prove immune function (1, 2). In recent years, it has become clear 1R domain–containing adapter inducing IFN-b (TRIF) bind to that the importance of zinc for immunity is based, at least in part, TLR4, resulting in a delayed NF-kB signal and activation of the on a function in in cells of the immune system IFN regulatory factor 3 (IRF3) (11). Signaling via TRIF and IRF3 (3). Intracellular zinc concentrations are tightly regulated by zinc triggers the secretion of IFN-b (12). IFN-b binds to the type 1 transporters and zinc signals (4). The latter consists of a change in IFNR, which, in turn, activates a JAK-STAT pathway inducing the the cytoplasmic concentration of free zinc ions (Zn2+), which have expression of surface molecules required for the interaction with already been observed in response to several stimuli and in many T cells, such as CD40, CD80, and CD86 (13, 14). Together, the ac- different types of immune cells. In monocytes, MCP-1 (5), PMA tivation of NF-kB and JAK-STAT also induces the expression of (6), TNF-a, insulin, and the TLR ligands Pam3CSK4 and LPS (7) inducible nitric monoxide synthase (iNOS) (15–17). iNOS activity by guest on October 6, 2021 induce an intracellular increase of the cytosolic free Zn2+ con- is mainly regulated through gene expression, and its product NO is centration. an important mediator of inflammation (18). Functions of iNOS- TLRs are a subgroup of the pattern-recognition receptors, which derived NO are the elimination of microbes, parasites, and cancer detect conserved molecular structures of pathogens that are used as cells (18, 19). danger signals by certain immune cells (8). Ligand binding to the Still, it is poorly understood how the TLR-induced inflammatory ectodomain of TLRs induces dimerization, bringing their cyto- and antiviral immune responses are balanced. Zinc signals affect plasmic Toll/IL-1R domains together, resulting in the recruitment the TLR4-induced production of MyD88-dependent inflammatory of intracellular adapter proteins and initiation of downstream cytokines (7), and the aim of this study is to investigate whether signaling events (9). TLR4 initially binds the adaptors TIRAP and intracellular Zn2+ levels also affect the TLR4-induced activation MyD88, and causes phosphorylation of MAPKs and early acti- of the TRIF pathway, in particular, the release of the inflammatory vation of NF-kB. The zinc signal is required for preventing de- mediator NO. phosphorylation of MAPKs p38, MEK1/2, and ERK1/2, and for Materials and Methods Cell culture Institute of Immunology, Medical Faculty, RWTH Aachen University, 52074 Aachen, All cell lines were cultured at 37˚C in a humidified 5% CO2 atmosphere. Germany RAW 264.7 murine macrophages, L929 fibroblasts, and 7FD3 hybridoma Received for publication May 13, 2013. Accepted for publication June 13, 2013. cells were grown in RPMI 1640 (Lonza, Verviers, Belgium) containing This work was supported by German Research Council (Deutsche Forschungsge- 10% FCS (heat inactivated for 30 min at 56˚C; PAA, Co¨lbe, Germany), 2 meinschaft) Grant HA4318/3 (to H.H.). mM L-glutamine, 100 U/ml potassium penicillin, and 100 U/ml strepto- mycin sulfate (all from Lonza). 3T3 and 293T cells were grown in DMEM Address correspondence and reprint requests to Prof. Hajo Haase, Institute of Immu- nology, Medical Faculty, RWTH Aachen University, Pauwelsstrasse 30, 52074 Aa- (Lonza) supplemented as described earlier. Zinc-deficient medium was chen, Germany. E-mail address: [email protected] obtained by treatment with CHELEX 100 ion exchange resin (Sigma- Aldrich, St. Louis, MO) as described previously (20). The online version of this article contains supplemental material. Abbreviations used in this article: BMM, bone marrow–derived macrophage; iNOS, Bone marrow–derived macrophages inducible nitric monoxide synthase; IRF3, IFN regulatory factor 3; PTP, protein tyrosine phosphatase; TPEN, N,N,N’,N’-Tetrakis(2-pyridylmethyl)ethylenediamine; Bone marrow was isolated from the femurs of C57BL/6 mice. A total of 5 TRIF, Toll/IL-1R domain–containing adapter inducing IFN-b. 3 3 10 cells/ml was seeded in 20 ml RPMI 1640 containing 18% heat- inactivated FCS, 2 mM L-glutamine, 100 U/ml potassium penicillin, 100 Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 U/ml streptomycin sulfate, 1 mM nonessential amino acids (Lonza), and www.jimmunol.org/cgi/doi/10.4049/jimmunol.1301261 The Journal of Immunology 1809

20% sterile-filtered L929 cell culture supernatant for a total of 8 d. On LumiGlo Reagent (Cell Signaling Technology) on an LAS-3000 (Fujifilm day 4, 10 ml fresh culture medium was added. On day 7, the medium was Lifescience, Du¨sseldorf, Germany). Densitometric quantification was per- completely replaced. On the next day, experiments were performed. formed with Adobe Photoshop Elements. Production of IFN-b–containing cell-free culture supernatant Nuclear extraction for Western blot samples The supernatants of 3T3 and 293T cells stably transfected to produce mouse A total of 5 3 106 cells was thoroughly suspended in 2 ml ice-cold buffer IFN-b were collected 4 d after seeding and were centrifuged for 10 min at (10 mM Tris-HCl, pH 7.4, containing 15 mM NaCl, 60 mM KCl, 0.5% 300 3 g. Supernatants were sterile-filtered with a 0.22-mm filter (Millex- Igepal CA-630, 1 mM EDTA, 0.1 mM EGTA) and lysed at 4˚C for 5 min. GV; Millipore, Billerica, MA), and aliquots were stored at 280˚C until Nuclei were pelleted by centrifugation (600 3 g, 4˚C, 5 min), taken up in use. The activities of the IFN-b–containing cell supernatants were deter- 100-ml Western blot sample buffer, and treated as described earlier for mined by an IFN-b bioassay with the L929 cell line and the Encephalo- Western blot samples. myocarditis virus. For all experiments, 2000 U/ml IFN-b was applied. RT-PCR Purification of monoclonal anti–IFN-b Abs The mRNA of 5 3 105 cells was isolated after lysis in 1 ml Tri Reagent 7FD3 hybridoma cells produce rat IgG1 anti-mouse IFN-b Abs. The cells (Ambion, Life Technologies, Carlsbad, CA) and transcribed into cDNA were grown for 3–4 d under normal culture conditions until confluency and with the qScript cDNA Synthesis Kit (Quanta Biosciences, Darmstadt, cultured for 3 more days in serum-free medium (21). Supernatants were Germany) according to the manufacturers’ instructions. Quantitative real- collected, sterile-filtered, and concentrated (Minicon B15 clinical sample time PCR was performed on an ABI PRISM 7000 or a Step-One plus Real- concentrator; Millipore). The Ab was purified using an IgG Purification Kit Time PCR System (both from Applied Biosystems, Darmstadt, Germany) (Pierce, Thermo Scientific, Rockford, IL) following the manufacturer’s with the oligonucleotide sequences shown in Table I. The components of instructions and quantified using an IgG-Standard with a Bradford assay each PCR sample (final 25 ml) were 8 mldH2O, 12.5 ml Power SYBR- (Bio-Rad Laboratories). The purified Ab solution was sterile-filtered with Green PCR MasterMix (Applied Biosystems), 1.25 ml forward and reverse a 0.22-mm filter (Millex-GV; Millipore), and 0.01–1 mg/ml of the Abs Downloaded from primer each (2 mM), and 2 ml of the respective cDNA (50 ng/ml) or dH2O applied as depicted in Fig. 5C. as a negative control. The real-time PCR was performed in duplicate with 2+ the following parameters for iNOS and STAT1: 95˚C for 15 min followed Zn measurements with fluorescent probes by 40 cycles of 95˚C for 30 s and 58˚C for 30 s; and for the other oli- Cells were grown on a 96-well plate and incubated in loading buffer (25 mM gonucleotides: 95˚C for 15 min followed by 40 cycles of 95˚C for 30 s, HEPES [pH 7.35], 120 mM NaCl, 5.4 mM KCl, 5 mM glucose, 1.3 mM 56˚C for 1 min, and 72˚C for 1 min. For quantification, the comparative cycle threshold method (DDCT) was used, normalizing the results to the CaCl2, 1 mM MgCl2, 1 mM NaH2PO4, 0.3% BSA) for 30 min with 1 mM

FluoZin-3-AM (Invitrogen, Eugene, OR) at 37˚C. Subsequently, cells were housekeeping gene b-actin. http://www.jimmunol.org/ washed twice with measurement buffer (incubation buffer without BSA). The resulting fluorescence was recorded on a Tecan Ultra 384 fluorescence Griess assay well plate reader (Tecan, Crailsheim, Germany) using an excitation Griess assay was performed modified after Griess (22). A total of 50 ml cell wavelength of 485 nm and measuring the emission at 535 nm. Intracellular 2+ supernatant was incubated for 10 min with 50 ml of a 1% Sulfanilamide free Zn concentrations were calculated as previously described (6), using (Fluka, Sigma-Aldrich) solution in 5% phosphoric acid. A total of 50 mlof 50 mM N,N,N’,N’-Tetrakis(2-pyridylmethyl)ethylenediamine (TPEN; Sigma- a 0.1% N-(1-naphtyl)ethylenediamine dihydrochloride (Riedel-de-Hae¨n; Aldrich) to determine minimal and 100 mMZnSO4/50 mM pyrithione to Sigma-Aldrich) solution was added and incubated for 10 min. All samples determine maximal fluorescence, respectively. TLR ligands were pur- were analyzed in duplicate. Nitrite production was determined by mea- chased from Invivogen (San Diego, CA). suring the absorption at 520 nm in a Sunrise well plate reader (Tecan) and

Western blotting quantified with a standard curve from 0.78 to 50 mM nitrite. by guest on October 6, 2021 A total of 5 3 105 cells was collected by centrifugation, lysed by soni- Neutral red assay m cation in 100 l sample buffer (62.5 mM Tris-HCl [pH 6.8], 2% [w/v] Cellular viability was measured by neutral red uptake. RAW 264.7 cells SDS, 27% [v/v] glycerol, 0.1% [v/v] 2-ME, 0.01% [w/v] bromophenol 3 5 3 (5 10 cells/ml) were grown for 24 h and treated for further 24 h with blue, 1 mM Na3VO4), and heated for 3 min at 95˚C. An equivalent of 1 LPS and TPEN. Cells were loaded for 3 h with neutral red (final concen- 105 cells/lane was separated on 10 or 14% (H3) polyacrylamide gels at 150 tration, 55 mM), washed with PBS, lysed in a mixture of ethanol/H2O/acetic V and blotted to nitrocellulose membranes. Uniform loading of gels was acid (50:49:1), and neutral red uptake was quantified by its absorption confirmed by staining with Ponceau S. After destaining, membranes were at 540 nm in a Sunrise well plate reader (Tecan). blocked overnight with TBST (20 mM Tris-HCl [pH 7.6], 136 mM NaCl, 0.1% [v/v] Tween 20) containing 5% fat-free dry milk and incubated with Flow cytometric measurement of endocytosis the primary Abs at 1/1000 (anti-H3 1/20,000) dilution in TBST containing 5% BSA for 3 h. All primary Abs were purchased from Cell Signaling TLR4 endocytosis was measured by flow cytometry as described previously Technology with the exception of anti-H3 (Abcam, Cambridge, U.K.), (10). After incubation, RAW 264.7 cells were treated with anti-CD16/ anti-TRAF3, and anti–NF-kB (Santa Cruz Biotechnology). Membranes CD32 blocking Ab for 5 min in ice-cold PBS. Anti–TLR4-PE (clone were washed three times with TBST and incubated for 1 h with goat anti- MTS510) or PE-conjugated rat IgG2a,k as isotype-matched control were rabbit–HRP (Cell Signaling Technology) followed by detection with added (all Abs were from BD Pharmingen, Heidelberg, Germany). Abs

Table I. Oligonucleotide sequences used for real-time PCR

Gene Forward Primer Reverse Primer iNOS 59-GGCAGCCTGTGAGACCTTTG-39 59-GCATTGGAAGTGAAGCGTTTC-39 b-Actin (48) 59-CGTGCGTGACATCAAAGAGA-39 59-CCATACCCAAGAAGGAAGGC-39 Rantes 59-CTCACCATCATCCTCACTGC-39 59-TCGAGTGACAAACACGACTG-39 IP-10 59-TGCCGTCATTTTCTGCCTC-39 59-TCGTGGCAATGATCTCAACA-39 MCP-1 59-TCAGCCAGATGCAGTTAACG-39 59-TCTGGACCCATTCCTTCTTG-39 IFN-a 59-AGTGATGAGCTACTGGTCAAC-39 59-TGATGCTGTGGAAGTATATCCT-39 IFN-b 59-CTCTCCATCAACTATAAGCA-39 59-CTTTCAAATGCAGTAGATTC-39 CD40 59-CCTTCTTCTTCTCACTAGAAA-39 59-ATTACAGATTTATTTAGCCAGT-39 CD80 59-CTCTACGACTTCACAATGTT-39 59-TTGATAGTCTCTCTGTCAGC-39 CD86 59-CTCTACGACTTCACAATGTT-39 59-TTGATAGTCTCTCTGTCAGC-39 IL-1b 59-AGTCAGGGTTCTTCCGTTAG-39 59-ACCTGTAAGGAGTGACACCA-39 IL-6 59-AGTTTTCCACCGTAAAGTGT-39 59-CGTCTCAAGGGGTTGACCAT-39 IL-10 59-GGATTTAGAGACTTGCTCTT-39 59-GTTTTAGCTTTTCATTTTGA-39 STAT1 59-TGCCGTGGAGCCCTACACGA-39 59-GCTCCATCGGTTCTGGTGCTTCC-39 1810 REGULATION OF TLR-DEPENDENT MYD88 AND TRIF SIGNALING BY ZINC and cells were incubated for 20 min at 4˚C in the dark and washed twice vating RAW 264.7 cells for 3 d either in CHELEX-treated medium with ice-cold PBS. Fluorescence was measured with a FACScan (Becton (zinc depleted) or CHELEX-treated medium reconstituted with 10 Dickinson, Heidelberg, Germany). mMZn2+ (zinc sufficient). The NO release of RAW 264.7 cells Statistical analysis cultivated in zinc-depleted medium is markedly higher than in cells 2+ Statistical significance was calculated by ANOVA using GraphPad Prism cultivated in zinc-sufficient medium. Furthermore, addition of Zn software (version 5.01). For the experiments with two variables (Fig. 1B–F), and pyrithione simultaneous to LPS stimulation diminishes NO two-way ANOVAwas used; the remaining data were analyzed by one-way production in cells grown in zinc-depleted medium (Fig. 1F). , ANOVA. A p value 0.05 was considered statistically significant. iNOS activity is mainly regulated on the transcriptional level. Accordingly, iNOS transcription in response to LPS stimulation of Results RAW 264.7 cells is elevated in the presence of TPEN (Fig. 2A). Role of Zn2+ in NO release by macrophages Notably, these conditions do not affect cellular viability (Sup- In the mouse macrophage cell line RAW 264.7, the fluorescent plemental Fig. 1B), and iNOS transcription is unaffected by TPEN probe FluoZin-3 detects an elevation of free intracellular Zn2+ in alone (Supplemental Fig. 1C). Furthermore, iNOS protein expres- response to LPS (Fig. 1A). Accordingly, an impact of Zn2+ on the sion after LPS stimulation is also increased by TPEN in a con- TLR4-induced release of NO is investigated. To this end, the centration-dependent manner (Fig. 2B). Altogether, the LPS-induced concentration of nitrite, a degradation product of NO, is taken as NO release is increased by Zn2+ chelation through elevated iNOS a measure for production of the short-lived NO. Addition of the expression. 2+ membrane-permeable Zn chelator TPEN before stimulation of 2+ primary murine bone marrow–derived macrophages (BMMs) with Impact of Zn chelation on TLR4 signaling LPS (Fig. 1B) and of RAW 264.7 cells with LPS or E. coli (Fig. TPEN has a differential impact on the mRNA expression of LPS- Downloaded from 1C, 1D) leads to a concentration-dependent increase of nitrite in inducible genes. In contrast with iNOS, the LPS-induced tran- the cell culture supernatants. Incubation of RAW 264.7 cells with scription of IL-1b, IL-6, and IL-10 is significantly reduced by equimolar concentrations of Zn2+ in addition to TPEN abolishes incubation of RAW 264.7 cells with the chelator (Fig. 3). On the the effect of the chelator (Fig. 1C), confirming that TPEN aug- other hand, transcription of type 1 IFNs is upregulated by Zn2+ ments NO production by chelation of Zn2+. In contrast, increasing chelation, in particular, IFN-b (Fig. 3). Rantes transcription is the intracellular Zn2+ concentration by addition of Zn2+ together only reduced by 1.25 mM TPEN, the lowest applied TPEN con- http://www.jimmunol.org/ with the ionophore pyrithione is without effect (Fig. 1D). centration, and there is no effect of TPEN on MCP-1 and IP-10 To exclude a contribution of the other NO synthases, eNOS and (Fig. 3). The transcription of the cell-surface protein CD40 is nNOS, the selective iNOS inhibitors 1400W and SMT are used. slightly increased, but the effect is not statistically significant. Both inhibitors completely abrogate LPS-induced NO production, CD80andCD86transcriptionissignificantly elevated (Fig. 3). regardless of the presence of TPEN (Fig. 1E, Supplemental Fig. The gene transcription pattern suggests differential effects of Zn2+ 1A). This confirms that the entire NO production in response to chelation on MyD88 and TRIF-dependent TLR4 signaling, reducing LPS is mediated exclusively by iNOS. Next, the effect of zinc MyD88 signaling whereas promoting the TRIF-dependent pathway. deficiency was investigated independently from TPEN by culti- Activation of the two pathways is balanced through endocytosis of by guest on October 6, 2021

FIGURE 1. Impact of free intracellular Zn2+ on nitrite production. (A) RAW 264.7 cells were loaded with FluoZin-3. After 10 min measuring the baseline fluorescence, LPS (10 mg/ml) was added (arrow) and the fluorescence measured for additional 50 min. (B–F) Nitrite concentrations in the supernatants were determined with the Griess assay after 24-h incubation under the following conditions. (B) BMMs were incubated as indicated with TPEN (5 min before LPS) and LPS. (C) RAW 264.7 cells were preincubated for 5 min with different concentrations of TPEN or TPEN and Zn2+ (1:1 molar ratio) and stimulated with LPS. (D) RAW 264.7 cells were preincubated for 5 min with TPEN or Zn2+/pyrithione (Z/P, both 1.5 mM) and stimulated with heat-killed E. coli.(E) RAW 264.7 cells were preincubated with TPEN for 5 min and stimulated with LPS (100 ng/ml) in the presence of the iNOS inhibitor 1400W. (F) RAW 264.7 cells were grown for 3 d either in zinc-depleted medium (CHELEX) alone or reconstituted with 10 mMZn2+ (CHELEX/Zn). Subsequently, cells were stimulated as indicated with LPS. Zn2+/pyrithione (both 1.5 mM) were added to cells grown in zinc-depleted medium (CHELEX+Z/P) simultaneously to LPS. All data are shown as means 6 SEM of at least n = 3 independent experiments. There was a statistically significant impact on nitrite production by TPEN (B–D; p , 0.001) and by CHELEX (F; p = 0.0011; two-way ANOVA). The Journal of Immunology 1811

TRAF3isubiquitinatedonLys48 for MyD88 signaling (with sub- sequent degradation), whereas it is ubiquitinated on Lys63 for TRIF signaling (23). No TRAF3 degradation is induced in BMMs treated from 15 min to 3 h with LPS, independently of TPEN preincubation (Fig. 4B). A time-dependent comparison between the activation of the transcription factors NF-kB and IRF3 was made in whole cells and nuclear extracts (Fig. 4C, 4D). The MyD88-dependent early nuclear translocation of NF-kB 30 min and 1 h after LPS stimulation is reduced in the presence of TPEN. In contrast, the TRIF-dependent delayed NF-kB activation after 2 and 3 h is not diminished by TPEN. IRF3 phosphorylation is detected in the whole-cell lysate starting at 1 h (Fig. 4D). After 2 and 3 h, phospho-IRF3 is found in the nucleus and in whole cells, but only nuclear phospho-IRF3 is increased by TPEN (Fig. 4C, 4D). Hence, Zn2+ chelation upregu- lates TRIF signaling upstream of the transcription factor IRF3 and leads to increased nuclear accumulation of phospho-IRF3. To further evaluate the time dependence of Zn2+ chelation for the enhancement of NO release, we added TPEN at different time points relative to LPS, ranging from 5 min prior (as in the previous Downloaded from FIGURE 2. Effect of TPEN on iNOS induction. (A) iNOS mRNA in experiments) to 8 h after addition of the TLR ligand. The addition RAW 264.7 cells after 5-min preincubation with TPEN and 4-h stimulation of TPEN 5 min before LPS stimulation doubles NO release in with LPS (100 ng/ml). Results are normalized to LPS-treated control and comparison with LPS alone, but the effect is even more pro- represent means + SEM of n = 7 independent experiments. (B) iNOS nounced at 30 min to 1 h after LPS stimulation (Fig. 4E). Starting protein in RAW 264.7 cells was measured by Western blotting after pre- 2 h after LPS stimulation, the effect of TPEN addition decreases incubation with TPEN and 10-h stimulation with 100 ng/ml LPS. Results slightly, but up to 8 h after LPS stimulation still increases NO http://www.jimmunol.org/ show means + SEM of densitometric quantifications (upper panel) and one release ∼2.5 times compared with LPS alone. By this time, the representative (lower panel) from n = 8 independent experiments. (A and B LPS-induced zinc signal has already returned to baseline levels ) Data were analyzed by repeated-measures ANOVA and Tukey post hoc 2+ test. Significantly different means do not share the same letters. (Fig. 4F). Hence, the effect of Zn chelation is most effective at the time when the TRIF pathway is activated, and basal Zn2+ levels, not the initial zinc signal, alter NO release. b 2+ the TLR4R complex (10). However, the LPS-induced endocytosis of Involvement of IFN- in the impact of Zn chelation on iNOS TLR4 is not affected by TPEN (Fig. 4A). TRAF3-ubiquitination is The main function of IRF3 in monocytes is the induction of IFN-b. another mode of balancing MyD88 and TRIF-dependent signaling. Because IFN-b transcription is elevated by Zn2+ chelation, the by guest on October 6, 2021

FIGURE 3. Effect of TPEN on the transcription of LPS-induced genes. The effects of 5-min pre- incubation with TPEN on mRNA expression after stimulation of RAW 264.7 cells with LPS (100 ng/ ml, 4 h) was analyzed for IL-1b, IL-6, IL-10, IFN-a, IFN-b, Rantes, MCP-1, IP-10, CD40, CD80, and CD86. Results are normalized to the LPS-treated control and represent means + SEM of n = 5 inde- pendent experiments. Data were analyzed by re- peated-measures ANOVA and Tukey post hoc test. Significantly different means do not share the same letters. 1812 REGULATION OF TLR-DEPENDENT MYD88 AND TRIF SIGNALING BY ZINC Downloaded from http://www.jimmunol.org/ by guest on October 6, 2021

FIGURE 4. Impact of Zn2+ chelation on TLR4 signaling. (A) Surface expression of TLR4 after stimulation with LPS (100 ng/ml) alone or in the presence of TPEN (1.75 mM, 5 min before LPS) was measured by flow cytometry. (B) Western blot of TRAF3 protein levels in BMMs after stimulation with LPS (100 ng/ml) alone or in the presence of TPEN (1.75 mM, 5 min before LPS). (C and D) RAW 264.7 cells were stimulated for the indicated times with LPS (100 ng/ml) and TPEN (5 min before LPS). Western blots for NF-kB, phosphorylated (Ser396) and total IRF3 were performed with nuclear extracts (C) and whole cells (D), using histone H3 and b-actin as respective housekeeping genes. (E) RAW 264.7 cells were incubated with LPS (100 ng/ml), and TPEN (1.75 mM) was added at different time points relative to LPS as indicated. A Griess assay was performed with the supernatant 24 h after addition of LPS. (F) Intracellular Zn2+ was measured as described for Fig. 1A. After the first hour, measurements were made in 30-min intervals. Results are shown as means 6 SEM of n = 3 independent experiments (A, E, F) or one representative experiment of n =3(B–D). impact of supplementation with additional IFN-b on NO release Stimulation increases STAT1 expression, which is abrogated by was examined. Stimulation with a high amount of IFN-b (2000 U/ preincubation with TPEN (Fig. 5E, 5F, Supplemental Fig. 2D). ml) alone does not induce the release of NO but augments LPS- Together, Zn2+ chelation has no effect on IFN-b–induced STAT1 induced NO release in a manner similar to TPEN. Notably, TPEN phosphorylation while reducing its transcription. does not cause a further increase of NO production (Fig. 5A, 2+ Supplemental Fig. 2A) and iNOS protein levels (Fig. 5B; Sup- Role of Zn in NO release in response to other receptors plemental Fig. 2B) in the presence of IFN-b. Purified rat IgG1 Abs To elucidate whether zinc signals are relevant for signaling by other against murine IFN-b cause a concentration-dependent reduction TLRs in addition to TLR4, we tested a panel of different ligands. of LPS-induced NO release by RAW 264.7 cells treated with LPS Stimulation of RAW 264.7 cells with pam3CSK4 (TLR1/2 ligand), and LPS+TPEN (Fig. 5C). Together, these results imply that Zn2+ heat-killed Listeria monocytogenes (TLR2 ligand), flagellin (TLR5 chelation augments the NO release through increased IFN-b levels. ligand), FSL-1 (TLR6/2 ligand), imiquimod and ssRNA40 (TLR7 IFN-b activates a JAK-STAT pathway; hence the impact of Zn2+ ligands), and ODN1826 (TLR9 ligand) induces zinc signals (Fig. 6). chelation on STAT1 phosphorylation was investigated. RAW 264.7 The only examined TLR ligand unable to induce a zinc signal cells were preincubated with TPEN and stimulated with IFN-b. is poly(I:C) (TLR3 ligand; Fig. 6). Comparable observations were Western blot analysis of STAT1 phosphorylation shows no effect of made using human primary monocytes (Supplemental Fig. 3). The TPEN on Tyr701 and Ser727 phosphorylation. However, there is a inability of poly(I:C) to induce a zinc signal was confirmed in slight decrease in total STAT1 protein in TPEN-preincubated and RAW 264.7 with the fluorescent probe ZinPyr1 (Supplemental IFN-b–stimulated samples (Fig. 5D, Supplemental Fig. 2C). Fig. 4A), and biological activity of the poly(I:C) in RAW 264.7 Consequently, transcription of STAT1 mRNA after preincubation cells was confirmed by measuring TNF-a secretion (Supplemental with TPEN and stimulation with LPS or IFN-b was measured. Fig. 4B). The Journal of Immunology 1813

FIGURE 5. Role of IFN-b in iNOS induction. (A) Nitrite was quantified in supernatants from RAW 264.7 cells preincubated with different concentrations of TPEN or Zn2+/pyrithione (Z/P, both 1.5 mM) and stim- ulated for 24 h with LPS (100 ng/ml) and IFN-b (2000 U/ml; 293T cell-free culture supernatant [CFCS]) or unconditioned control medium. (B) RAW 264.7 cells were incubated for 10 h as described for (A), and iNOS expression was determined by Western blotting. (C) Nitrite was quantified in supernatants from RAW 264.7 cells preincubated with TPEN (1.75 mM) and stimulated for 24 h with LPS (100 ng/ml) in the presence of purified rat IgG1 Abs against murine IFN-b.(D) RAW 264.7 cells were incubated for 3 h with TPEN (1.75 mM) and additionally for 30 min with IFN-b (2000 U/ml; 293T CFCS) or unconditioned control medium. Phosphoryla- tion of STAT1 at Tyr701 and Ser727 was determined by Western blotting. (E and F) STAT1 mRNA was deter- mined in RAW264.7 cells. (E) Five-minute preincubation with TPEN was followed by 4-h stimulation with LPS Downloaded from (100 ng/ml). (F) Five-minute preincubation with TPEN (1.75 mM) was followed by 4-h stimulation with IFN-b (2000 U/ml; 293T CFCS) or unconditioned control me- dium. (A, C, E,andF) Results are shown as means + SEM of n = 3 or (B, D) one representative of n = 3 independent experiments. (E,andF) Data were analyzed by repeated- measures ANOVA and Tukey post hoc test. Significantly http://www.jimmunol.org/ different means do not share the same letters.

Further investigations determined the effect of Zn2+ chelation on (Fig. 7C). Comparable with the observations with TLR4, the NO release induced by another TLR that triggered a zinc signal, levels of IFN-a, IFN-b, CD40, CD80, and CD86 mRNA are in- TLR7, and by TLR3, which is inactive with respect to zinc sig- creased, although the effect is only statistically significant for naling. Comparable with LPS, the TLR7 ligand imiquimod IFN-b, CD80, and CD86 (Fig. 7C). The effect of TPEN on TLR3- induces the release of NO, which can be further elevated by TPEN induced NO release and IFN-b mRNA transcription was also by guest on October 6, 2021 (Fig. 7A). iNOS protein expression is also increased by incubation examined. NO release in response to poly(I:C) increases in a con- of imiquimod-stimulated RAW 264.7 cells with TPEN (Fig. 7B). centration-dependent manner in TPEN-treated samples (Fig. 7D). Similar to TLR4, Zn2+ chelation affected the transcription of In addition, IFN-b transcription is elevated after preincubation with several genes. Whereas the mRNA levels of IL-1b and IL-6 are TPEN and stimulation with poly(I:C) (Fig. 7E). Similar to the effect unaltered by TPEN, there is a significant decrease of IL-10 mRNA on TLR4 signaling, increasing the intracellular free Zn2+ concen-

FIGURE 6. TLR-induced zinc signals in RAW 264.7 cells. Intracellular Zn2+ was measured as de- scribed for Fig. 1A. RAW 264.7 cells were stimulated with Pam3CSK4 (1 mg/ml), heat-killed L. monocytogenes (HKLM, 107/ml), poly(I:C) (10 mg/ml), Salmonella typhimurium flagellin (1 mg/ml), FSL-1 (1 mg/ml), imi- quimod (10 mg/ml), ssRNA40 (1 mg/ml), or ODN1826 (10 mg/ml). Data are shown as means 6 SEM of at least n = 3 independent experiments. 1814 REGULATION OF TLR-DEPENDENT MYD88 AND TRIF SIGNALING BY ZINC Downloaded from http://www.jimmunol.org/

FIGURE 8. Role of Zn2+ in IFN-g–induced NO release. (A) Intracellular Zn2+ was measured as described for Fig. 1A. RAW 264.7 cells were stim- ulated with LPS (1 mg/ml) or IFN-g (500 U/ml). (B) Supernatant nitrite by guest on October 6, 2021 concentrations of RAW264.7 cells were determined with the Griess assay after 24-h incubation with TPEN (5 min before LPS/IFN), IFN-g (500 U/ ml), or LPS (100 ng/ml). Statistically significant differences to the respec- tive controls are indicated (repeated-measures ANOVAwith Dunnett’s post hoc test; **p , 0.01). (C) Survival of RAW 264.7 cells incubated as indi- cated with TPEN, IFN-g (500 U/ml), or LPS (100 ng/ml) was determined by neutral red assay. All data are shown as means from at least n = 3 inde- pendent experiments 6 SEM.

tration by addition of Zn2+ and pyrithione does not alter TLR3- mediated NO release or IFN-b transcription (Fig. 7D, 7E). IFN-g induces iNOS expression independently from IFN-b (24). In contrast with LPS, IFN-g does not trigger a zinc signal FIGURE 7. Impact of Zn2+ chelation on TLR3- and TLR7-induced iNOS induction. (A) RAW 264.7 cells were preincubated with TPEN (5 min) before (Fig. 8A) and IFN-g–mediated NO production is not augmented addition of imiquimod (1 mg/ml, 24 h), followed by nitrite determination with by TPEN (Fig. 8B). Notably, the absence of TPEN-induced ele- a Griess assay. (B) iNOS protein expression in RAW 264.7 cells was determined vation of iNOS activity is not due to toxicity of the chelator in by Western blotting after 5-min preincubation with TPEN and stimulation for combination with IFN-g (Fig. 8C). Consequently, regulation by 24 h with imiquimod (1 mg/ml) or LPS (100 ng/ml). (C)mRNAlevelsofIL-1b, free Zn2+ is not a general feature of iNOS expression but is spe- IL-6,IL-10,IFN-a,IFN-b, CD40, CD80, and CD86 in RAW 264.7 cells treated cific for the TRIF/IRF/IFN-b pathway. with TPEN (5 min before imiquimod) and imiquimod (1 mg/ml, 4 h). Results are normalized to the imiquimod-treated controls. (D) RAW 264.7 cells were Discussion 2+ stimulated with poly(I:C) (1 mg/ml) 5 min after addition of TPEN or Zn / Binding of LPS to TLR4 leads to recruitment of the adaptor pyrithione (Z/P, both 1.5 mM). Statistically significant differences to the un- proteins TIRAP and MyD88, resulting in the early activation of NF- treated control are indicated (repeated-measures ANOVAwith Dunnett post hoc kB, triggering mRNA expression of inflammatory cytokines (25). test; ***p , 0.001). (E)IFN-b mRNA levels were determined in RAW 264.7 Subsequently, the receptor complex is internalized and binds TRAM cells after 4-h stimulation with poly(I:C) (1 mg/ml) after addition of TPEN (1.75 mM), Zn2+ (Zn, 1.75 mM), and pyrithione (Pyr, 1.75 mM) as indicated. Results and TRIF, inducing the delayed activation of NF-kB and phos- are normalized to the untreated control. Results are shown as means + SEM (A, phorylation of IRF3. Phosphorylated IRF3 translocates into the C–E) or one representative blot (B)ofatleastn = 3 independent experiments. nucleus and induces the transcription of IFN-b (25). In turn, IFN-b Data in (A, C, E) were analyzed by repeated-measures ANOVA and Tukey post production is essential for LPS-induced iNOS expression (17), hoc test. Significantly different means do not share the same letters. because IFN-b activates the type I IFNR and the JAK-STAT The Journal of Immunology 1815 pathway in an autocrine and paracrine fashion (16). Together, NF- more importantly, TPEN was ineffective in the presence of excess kB and STAT1 induce iNOS transcription (15). IFN-b. Further downstream, a potential influence of Zn2+ chela- Activation of TLR4 also triggers a zinc signal. It supports MAPK tion on the IFN-b–dependent activation of the STAT pathway was signaling through inhibition of dephosphorylation, and is required examined. Treatment of cells with TPEN had no effect on IFN-b– for the phosphorylation of IKKa/b and for NF-kB–dependent mediated STAT1 phosphorylation. A comparable result has been production of inflammatory cytokines (7). This involvement of a observed for STAT5 phosphorylation in response to stimulation of zinc signal in MyD88 signaling is confirmed by this study. After the T cell line CTLL-2 with IL-2 (37), indicating that zinc signals TLR4 or TLR7 stimulation, the transcription of NF-kB–dependent might not be involved in JAK/STAT phosphorylation signaling by genes such as IL-1b, IL-6, and IL-10 decreased, at least to some receptors in general. In contrast, TPEN reduced STAT1 extent, when the zinc signal was chelated by TPEN. In addition, transcription in response to stimulation with LPS or IFN-b. This Zn2+ chelation reduced the MyD88-dependent nuclear transloca- finding will have no impact on the signaling events described in tion of NF-kB. Nevertheless, free intracellular Zn2+ is not an this article because the reported half-life for STAT1 is 16 h (38). It unspecific activator of all aspects of TLR signaling. The transcrip- could, however, affect long-term signaling via the IFN-bR, rep- tion of TRIF-dependent genes, such as IFN-b and the IFN-b–induced resenting a possible feedback mechanism to limit iNOS induction genes CD80 and CD86, is enhanced by Zn2+ chelation. Moreover, in response to prolonged exposure to LPS. This effect also could the Zn2+ chelator TPEN increases LPS-induced NO release through influence the subsequent activation of other signaling pathways upregulation of iNOS expression. Simultaneously, there is no dis- involving STAT1 phosphorylation and signals depending on the tincteffectofZn2+ chelation on the TRIF-dependent genes Rantes, level of unphosphorylated STAT1 protein for transcriptional ac- IP-10, and MCP-1. All three genes are thought to be mainly induced tivity (39). by TRIF signaling (26), but, possibly, other signaling pathways such Zn2+ frequently mediates its effects on signal transduction by Downloaded from as PI3K, NF-kB, and the MAPK ERK are also involved in TLR- inhibiting protein tyrosine phosphatases (PTPs) (40). PTPs known dependent induction of Rantes, MCP-1, and IP-10 (27, 28), ex- to be involved in TRIF signaling are SHP-2, SHIP1, and PTP1B plaining the differential outcome. (41–43). Nevertheless, all three PTPs are negative regulators of Signaling via several homodimers and heterodimers of TLR1-9 TRIF signaling, and activating them by reducing free Zn2+ could induces zinc signals, with TLR3 being the only exception. Nota- not explain augmentation of the TRIF pathway (41–43). More- bly, TLR3 is also the only TLR that does not activate the MyD88 over, the reduction of basal intracellular Zn2+ concentrations http://www.jimmunol.org/ pathway (29) but is capable of inducing IFN and iNOS expression modifies TRIF signaling, whereas most PTP inhibition constants via TRIF-dependent activation of NF-kB and IRF3 (30). Hence, not measured so far are higher than basal intracellular Zn2+ level; for 2+ only does the zinc signal selectively support MyD88-dependent example, PTP1B is inhibited by Zn with an IC50 of 17 nM (40), signaling, it only occurs in response to stimulation of TLRs trig- a concentration that is well above the basal intracellular Zn2+ gering that particular pathway. Nevertheless, LPS-induced zinc sig- measured in monocytes (6). Yet, there is significant divergence in nals occur in BMMs isolated from MyD88 knockout mice; therefore, the inhibition constants of different PTP for Zn2+. For the RPTPb, the adaptor molecule itself is not required for the induction of zinc an inhibition constant of 21 pM was recently shown, indicating signals (7). by guest on October 6, 2021 To identify the aspect(s) of TRIF signaling that are negatively regulated by Zn2+, we investigated several details of the pathway. Endocytosis of TLR4 was not altered by Zn2+ chelation, pointing toward an impact of Zn2+ that is located downstream of receptor internalization. Also, TRAF3 degradation was not involved. In contrast, an increase of nuclear, but not total cellular phospho- IRF3, was measured in response to Zn2+ chelation. This suggests that Zn2+ affects nuclear phosphorylation of IRF3, or nuclear translocation or retention of phospho-IRF3. Furthermore, TLR7- induced IFN-b and iNOS transcription also increases by Zn2+ chelation, but TLR7 signals via MyD88 and IRF7. IRF3 activation depends on phosphorylation by TBK1 and IKKε (31). The kinases are activated by the recruitment of TRAF3 to TRIF (32) and Lys63 ubiquitination of TRAF3 (23). IRF7, in contrast, is phosphorylated by the kinases IL-1R–associated kinase 1 and IKK-a (33). Unphos- phorylated IRF3 shuttles between cytoplasm and nucleus with nu- clear export being the predominant effect (34, 35). Transport pro- teins for IRF3 import are a subset of importin-a receptors, namely, Qip1 and KPNA3, which associate with importin-b receptors and the small GTPase Ran. Nuclear export is mediated by chromosome region maintenance/exportin 1 (34). Phosphorylated IRF3 and IRF7 FIGURE 9. Scheme of the role of Zn2+ in TLR4 signaling. TLR4 signals bind to the histone acetylases CREB-binding protein and p300 (35, via two main pathways involving the adaptor proteins MyD88 or TRIF. 36), and are localized in the nucleus through retention by CREB- MyD88-dependent signaling induces early activation of the transcription binding protein/p300 (35). Therefore, activation and nuclear trans- factor NF-kB, leading to the production of inflammatory cytokines, such as IL-1b and IL-6. TRIF-dependent signaling causes a delayed activation of location of IRF3, as well as IRF7, offer many possibilities for reg- 2+ NF-kB (not shown for reasons of clarity) and IRF3, resulting in the pro- ulation by Zn , which should be elucidated in more detail in future duction of IFN-b. In turn, IFN-b activates the type I IFNR and its corre- investigations. sponding JAK-STAT pathway in an autocrine and paracrine fashion, leading 2+ Enhancement of iNOS induction by Zn chelation also depends to the production of iNOS. Free intracellular Zn2+ has a differential impact on the IRF3-dependent increase in IFN-b production. NO release on both pathways, augmenting MyD88-dependent signaling whereas inhib- was almost completely abrogated by an IFN-b–specific Ab, and iting TRIF-mediated activation of IRF3. 1816 REGULATION OF TLR-DEPENDENT MYD88 AND TRIF SIGNALING BY ZINC that under normal conditions, RPTPb could exist predominantly 8. Medzhitov, R. 2009. Approaching the asymptote: 20 years later. Immunity 30: 2+ 766–775. in its Zn -bound inactive state, being activated by a reduction of 9. Botos, I., D. M. Segal, and D. R. Davies. 2011. The structural biology of Toll- 2+ basal Zn levels (44). Consequently, there may be one or more like receptors. Structure 19: 447–459. PTPs positively regulating TRIF signaling, which are controlled 10. Kagan, J. C., T. Su, T. Horng, A. Chow, S. Akira, and R. Medzhitov. 2008. 2+ TRAM couples endocytosis of Toll-like receptor 4 to the induction of interferon- by intracellular free Zn in a similar manner. It was suggested beta. Nat. Immunol. 9: 361–368. 2+ that Zn inhibits PTPs by binding to a catalytically active thiol in 11. Fitzgerald, K. A., D. C. Rowe, B. J. Barnes, D. R. Caffrey, A. Visintin, E. Latz, the active site (45). Other enzymes, such as deubiquitinases, also B. Monks, P. M. Pitha, and D. T. Golenbock. 2003. LPS-TLR4 signaling to IRF- 3/7 and NF-kappaB involves the toll adapters TRAM and TRIF. J. Exp. Med. have thiols in their catalytic center with a comparably low pKa, 198: 1043–1055. and hence similar chemical reactivity, which is illustrated by the 12. Monroe, K. M., S. M. McWhirter, and R. E. Vance. 2010. Induction of type I fact that both classes of enzymes can be inactivated reversibly by interferons by bacteria. Cell. Microbiol. 12: 881–890. 13. Hoebe, K., E. M. Janssen, S. O. Kim, L. Alexopoulou, R. A. Flavell, J. Han, and thiol oxidation (46). It is conceivable that these enzymes also have B. Beutler. 2003. Upregulation of costimulatory molecules induced by lipo- a similar reactivity toward Zn2+, making them potential molecular polysaccharide and double-stranded RNA occurs by Trif-dependent and Trif- 2+ independent pathways. Nat. Immunol. 4: 1223–1229. targets for an effect of Zn in addition to PTPs. 14. Lien, E., and D. T. Golenbock. 2003. Adjuvants and their signaling pathways: Dendritic cell maturation in response to LPS and monocytic dif- beyond TLRs. Nat. Immunol. 4: 1162–1164. ferentiation of myeloid precursors stimulated by calcitriol both 15. Farlik, M., B. Reutterer, C. Schindler, F. Greten, C. Vogl, M. Mu¨ller, and T. Decker. 2010. Nonconventional initiation complex assembly by STATand NF- depend on alterations of cellular zinc homeostasis, ultimately kappaB transcription factors regulates nitric oxide synthase expression. Immu- 2+ 2+ leadingtoreducedfreeZn (20, 47). This shows that Zn exerts nity 33: 25–34. regulatory functions not only through short-term spikes, but also 16. Gao, J. J., M. B. Filla, M. J. Fultz, S. N. Vogel, S. W. Russell, and W. J. Murphy. 2+ 1998. Autocrine/paracrine IFN-alphabeta mediates the lipopolysaccharide-induced via modulation of basal Zn levels. Both effects are combined in activation of transcription factor Stat1alpha in mouse macrophages: pivotal role of Downloaded from TLR4 signaling, balancing MyD88 and TRIF signaling (Fig. 9). Stat1alpha in induction of the inducible nitric oxide synthase gene. J. Immunol. After the initial increase in free Zn2+, facilitating MyD88 signaling, 161: 4803–4810. 17. Jacobs, A. T., and L. J. Ignarro. 2003. Cell density-enhanced expression of in- the concentration returns to basal values within a few hours, but ducible nitric oxide synthase in murine macrophages mediated by interferon- these are still sufficiently high to limit TRIF signaling. beta. Nitric Oxide 8: 222–230. The reaction of macrophages toward TLR stimulation is mul- 18. Kro¨ncke, K. D., K. Fehsel, C. Suschek, and V. Kolb-Bachofen. 2001. Inducible nitric oxide synthase-derived nitric oxide in gene regulation, cell death and cell tifaceted. An inflammatory immune response is triggered through survival. Int. Immunopharmacol. 1: 1407–1420. MyD88-dependent production of cytokines, such as IL-1b and 19.Hill,B.G.,B.P.Dranka,S.M.Bailey,J.R.Lancaster,Jr.,and http://www.jimmunol.org/ V. M. Darley-Usmar. 2010. What part of NO don’t you understand? Some IL-6. In addition, TLRs induce an antiviral immune response through answers to the cardinal questions in nitric oxide biology. J. Biol. Chem. TRIF-mediated release of IFN-b (48). TLR-induced NO release 285: 19699–19704. has various effects. NO is involved in antimicrobial, antiviral, and 20. Dubben, S., A. Ho¨nscheid, K. Winkler, L. Rink, and H. Haase. 2010. Cellular zinc homeostasis is a regulator in monocyte differentiation of HL-60 cells by 1 antitumor immune defense, but can also damage host tissue and alpha,25-dihydroxyvitamin D3. J. Leukoc. Biol. 87: 833–844. suppress beneficial as well as harmful immune responses (49). 21. Zawatzky, R., H. Wurmbaeck, W. Falk, and A. Homfeld. 1991. Endogenous This is of particular importance for the adequate activation of innate interferon specifically regulates Newcastle disease virus-induced cytokine gene expression in mouse macrophages. J. Virol. 65: 4839–4846. immunity. On the one hand, TLRs are necessary for the induction of 22. Griess, P. 1879. Bemerkungen zu der Abhandlung der HH. Weselsky und Ben- protective immunity against infection; on the other hand, an inap- edikt ueber einige Azoverbindungen. Berichte d. d. chem. Gesellschaft. 12: 426– by guest on October 6, 2021 propriate TLR response is associated with acute and chronic inflam- 428. 23. Tseng, P. H., A. Matsuzawa, W. Zhang, T. Mino, D. A. Vignali, and M. Karin. mation and systemic autoimmune diseases (48). This study identi- 2010. Different modes of ubiquitination of the adaptor TRAF3 selectively ac- fies free Zn2+ ions as a signal differentially affecting all of these tivate the expression of type I interferons and proinflammatory cytokines. Nat. Immunol. 11: 70–75. functions at multiple levels, hereby fine-tuning TLR signaling. 24. Schroder, K., P. J. Hertzog, T. Ravasi, and D. A. Hume. 2004. Interferon-gamma: an overview of signals, mechanisms and functions. J. Leukoc. Biol. 75: 163–189. Acknowledgments 25. Ostuni, R., I. Zanoni, and F. Granucci. 2010. Deciphering the complexity of Toll- like receptor signaling. Cell. Mol. Life Sci. 67: 4109–4134. We thank Prof. Dr. R. Zawatzky (Deutsches Krebsforschungszentrum Hei- 26. Yamamoto, M., S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, delberg, Heidelberg, Germany) for the 7FD3 hybridoma cell line and the O. Takeuchi, M. Sugiyama, M. Okabe, K. Takeda, and S. Akira. 2003. Role of 3T3 and 293T cell lines transfected to produce IFN-b. We also thank Silke adaptor TRIF in the MyD88-independent toll-like receptor signaling pathway. Hebel and Gabriela Engelhardt for technical assistance. Science 301: 640–643. 27. Mitchell, D., and C. Olive. 2010. Regulation of Toll-like receptor-induced chemokine production in murine dendritic cells by mitogen-activated protein Disclosures kinases. Mol. Immunol. 47: 2065–2073. The authors have no financial conflicts of interest. 28. Orita, T., K. Kimura, T. Nishida, and K. H. Sonoda. 2013. Cytokine and che- mokine secretion induced by poly(I:C) through NF-kB and phosphoinositide 3- kinase signaling pathways in human corneal fibroblasts. Curr. Eye Res. 38: 53–59. 29. Kawai, T., and S. Akira. 2007. TLR signaling. Semin. Immunol. 19: 24–32. References 30. Hiscott, J., T. L. Nguyen, M. Arguello, P. Nakhaei, and S. Paz. 2006. Manipu- 1. Prasad, A. S., F. W. Beck, B. Bao, J. T. Fitzgerald, D. C. Snell, J. D. Steinberg, lation of the nuclear factor-kappaB pathway and the innate immune response by and L. J. Cardozo. 2007. Zinc supplementation decreases incidence of infections viruses. Oncogene 25: 6844–6867. in the elderly: effect of zinc on generation of cytokines and oxidative stress. Am. 31. Fitzgerald, K. A., S. M. McWhirter, K. L. Faia, D. C. Rowe, E. Latz, J. Clin. Nutr. 85: 837–844. D. T. Golenbock, A. J. Coyle, S. M. Liao, and T. Maniatis. 2003. IKKepsilon and 2. Haase, H., and L. Rink. 2009. The immune system and the impact of zinc during TBK1 are essential components of the IRF3 signaling pathway. Nat. Immunol. 4: aging. Immun. Ageing 6: 9. 491–496. 3. Haase, H., and L. Rink. 2009. Functional significance of zinc-related signaling 32. Oganesyan, G., S. K. Saha, B. Guo, J. Q. He, A. Shahangian, B. Zarnegar, pathways in immune cells. Annu. Rev. Nutr. 29: 133–152. A. Perry, and G. Cheng. 2006. Critical role of TRAF3 in the Toll-like receptor- 4. Cousins, R. J., J. P. Liuzzi, and L. A. Lichten. 2006. Mammalian zinc transport, dependent and -independent antiviral response. Nature 439: 208–211. trafficking, and signals. J. Biol. Chem. 281: 24085–24089. 33. Blasius, A. L., and B. Beutler. 2010. Intracellular toll-like receptors. Immunity 5. Kojima, C., A. Kawakami, T. Takei, K. Nitta, and M. Yoshida. 2007. 32: 305–315. Angiotensin-converting enzyme inhibitor attenuates monocyte adhesion to vas- 34. Reich, N. C. 2002. Nuclear/cytoplasmic localization of IRFs in response to viral cular endothelium through modulation of intracellular zinc. J. Pharmacol. Exp. infection or interferon stimulation. J. Interferon Cytokine Res. 22: 103–109. Ther. 323: 855–860. 35. Kumar, K. P., K. M. McBride, B. K. Weaver, C. Dingwall, and N. C. Reich. 6. Haase, H., S. Hebel, G. Engelhardt, and L. Rink. 2006. Flow cytometric mea- 2000. Regulated nuclear-cytoplasmic localization of interferon regulatory factor surement of labile zinc in peripheral blood mononuclear cells. Anal. Biochem. 3, a subunit of double-stranded RNA-activated factor 1. Mol. Cell. Biol. 20: 352: 222–230. 4159–4168. 7. Haase, H., J. L. Ober-Blo¨baum, G. Engelhardt, S. Hebel, A. Heit, H. Heine, and 36. Yang, H., C. H. Lin, G. Ma, M. O. Baffi, and M. G. Wathelet. 2003. Interferon L. Rink. 2008. Zinc signals are essential for lipopolysaccharide-induced signal regulatory factor-7 synergizes with other transcription factors through multiple transduction in monocytes. J. Immunol. 181: 6491–6502. interactions with p300/CBP coactivators. J. Biol. Chem. 278: 15495–15504. The Journal of Immunology 1817

37. Kaltenberg, J., L. M. Plum, J. L. Ober-Blo¨baum, A. Ho¨nscheid, L. Rink, and 43. Xu, H., H. An, J. Hou, C. Han, P. Wang, Y. Yu, and X. Cao. 2008. Phosphatase H. Haase. 2010. Zinc signals promote IL-2-dependent proliferation of T cells. PTP1B negatively regulates MyD88- and TRIF-dependent proinflammatory Eur. J. Immunol. 40: 1496–1503. cytokine and type I interferon production in TLR-triggered macrophages. Mol. 38. Siewert, E., W. Mu¨ller-Esterl, R. Starr, P. C. Heinrich, and F. Schaper. 1999. Immunol. 45: 3545–3552. Different protein turnover of interleukin-6-type cytokine signalling components. 44. Wilson, M., C. Hogstrand, and W. Maret. 2012. Picomolar concentrations of free Eur. J. Biochem. 265: 251–257. zinc(II) ions regulate receptor protein-tyrosine phosphatase b activity. J. Biol. 39. Yang, J., and G. R. Stark. 2008. Roles of unphosphorylated STATs in signaling. Chem. 287: 9322–9326. Cell Res. 18: 443–451. 45. Kro¨ncke, K. D., and L. O. Klotz. 2009. Zinc fingers as biologic redox switches? 40. Haase, H., and W. Maret. 2003. Intracellular zinc fluctuations modulate protein Antioxid. Redox Signal. 11: 1015–1027. 46. Cotto-Rios, X. M., M. Be´ke´s, J. Chapman, B. Ueberheide, and T. T. Huang. tyrosine phosphatase activity in insulin/insulin-like growth factor-1 signaling. 2012. Deubiquitinases as a signaling target of oxidative stress. Cell Rep 2: Exp. Cell Res. 291: 289–298. 1475–1484. 41. An, H., W. Zhao, J. Hou, Y. Zhang, Y. Xie, Y. Zheng, H. Xu, C. Qian, J. Zhou, 47. Kitamura, H., H. Morikawa, H. Kamon, M. Iguchi, S. Hojyo, T. Fukada, Y. Yu, et al. 2006. SHP-2 phosphatase negatively regulates the TRIF adaptor S. Yamashita, T. Kaisho, S. Akira, M. Murakami, and T. Hirano. 2006. Toll-like protein-dependent type I interferon and proinflammatory cytokine production. receptor-mediated regulation of zinc homeostasis influences dendritic cell Immunity 25: 919–928. function. Nat. Immunol. 7: 971–977. 42. Gabhann, J. N., R. Higgs, K. Brennan, W. Thomas, J. E. Damen, N. Ben Larbi, 48. Kawai, T., and S. Akira. 2010. The role of pattern-recognition receptors in innate G. Krystal, and C. A. Jefferies. 2010. Absence of SHIP-1 results in constitutive immunity: update on Toll-like receptors. Nat. Immunol. 11: 373–384. phosphorylation of tank-binding kinase 1 and enhanced TLR3-dependent IFN- 49. MacMicking, J., Q. W. Xie, and C. Nathan. 1997. Nitric oxide and macrophage beta production. J. Immunol. 184: 2314–2320. function. Annu. Rev. Immunol. 15: 323–350. Downloaded from http://www.jimmunol.org/ by guest on October 6, 2021